Flim.ai

Flim is an AI powered food platform meant to reduce food waste at the most lowest levels - Households. Everyone thinks about canteens and cafe` but what about the lowest level, common household?

Comment

Idea

Explaination

  1. Problem - We waste millions of tonne of food annually and millions of people die due to hunger.
  2. Solution - We have made a application containing 3 parts.
    • Producers (Target: Household Food caretakers) - Example: Mom, Dad or Students who cook food.
    • Engine - Our application
    • Receivers (Target: Hungry People, Homeless people or Student tight on budget)
    • Our idea connects those producer to receiver.
  3. Phases:
    • Producer have excess food. They report it. (1 click!)
    • Recivers can see food. They request. (1 click)
    • Producers have to respond. (1 click)

Usage of AI

Use of AI: Buying human labour and team is expensive so we are using AI.

  • To detect FOOD.
  • To detect EXPIRY.
  • To detect DANGERs.

Issues

These can be prone to errors -> Fix in future = We can fine tune or give access to a reputable food dataset (RAG) and APIs.

What we did not implemented

  1. Delivery System

Story

I attended to whole webinar and got the PDF from DevPost. I thought creating help in Community wouldn't be too innovative; hundreds of people making it and all have the same core - A chatbot with system prompt. (Most probably) So to standout we chose environment.

In the directions I did not understood path b. So we ended up choosing Dir A.

First we saw the issue. (Food Waste) I used Gemini to think about some ideas. 10/9 times it said make a chatbot which can detect food usage (system predicts based on past data to predict the food courns eg. 14 sandwiches 5 pasta). It didn't seem good to me. Then I used my brain (for real) no ai. I thought we can go to the bare bones. Like Binary is the most low level. Houses are the most low level.

Instead of 500 canteens, we can target 5000 houses. The most major part of food wastage.

Doubt Pla

Built With

Share this project:

Updates